Optimum design of large-scale systems considering material nonlinearities and uncertainties. (15th October 2019)

• Record Type: Journal Article
• Authors: Giagopoulos, D.; Arailopoulos, A.; Chatziparasidis, I.; Sapidis, N.S.
• Is Part Of: Computers & structures. Volume 223(2019)
• Publication Date: 2019-10-15
• Subjects: Large-scale structures -- FE model updating -- Uncertainty quantification -- Structural dynamics -- Material nonlinearities; Structural engineering -- Data processing -- Periodicals; Electronic data processing -- Structures, Theory of -- Periodicals; 624.171
• DOI: 10.1016/j.compstruc.2019.106102
• Languages: English
• ISSNs: 0045-7949

Abstract

Highlights: Computational FE model updating framework for large scale dynamical systems. Covariance Matrix Adaptation Evolution Strategy (CMAES) optimization algorithm. Updating large-scale mechanical systems with strong material or structural nonlinearities. Numerical and experimental methodologies were applied to identify the parameters. Optimum design and validation. Abstract: A novel optimum design procedure for uncertainty quantification and validation, applicable in large-scale mechanical systems or industrial structures with material or structural nonlinearities is proposed in this work. Its implementation is presented through the research and optimal development of a full glass panoramic car elevator under real operational fail-safe loading scenario, including accurate dynamic analyses, placing emphasis on high fidelity FE models, uncertainty quantification and efficient handling of material nonlinearities. At first, a small-scale laboratory experimental arrangement of a single glass panel was examined, in order to develop high fidelity FE models of the laminated glass system and the point-support contact mechanism based on the level of dynamic excitation forces. Structural identification along with effective computational model updating and uncertainty quantification techniques were applied, in order to finely tune and estimate the parameters (material properties and damping ratios) of the numerical FE models and assess their uncertainties. The material nonlinearity of the natural rubber was examined at three forcing levels and excitation frequencies. Using maximum stress-strain pairs and updated moduli of elasticity a stress-strain curve, defining the nonlinear elastic behavior of the rubber was applied in the implicit nonlinear analysis of the full-scale elevator system. Based on the results of this analysis the elevator chassis was redesigned and optimized, in order to achieve minimum design stresses at the glazing components under emergency safety gear engagement. The stress levels developed on the optimal design of the elevator car were validated, in order to test the reliability of the applied method.